GLP-1 Microdosing and Anxiety: Does a Low Dose Calm Worry or Trigger Panic?
Some people say their GLP-1 quieted a restless mind. Others say it brought on panic. The research on GLP-1 microdosing and anxiety splits the same way, and most of the split comes down to how each study was built.
0.92
RISK RATIO FOR ANXIETY ACROSS 86 RANDOMISED GLP-1 TRIALS
0.62
HAZARD RATIO FOR WORSENING ANXIETY ON SEMAGLUTIDE IN SWEDEN
6.7 → 42.0
PANIC REPORTS PER 10,000 SEMAGLUTIDE REPORTS, BEFORE AND AFTER 2023
Anxiety and weight problems travel together. People with type 2 diabetes or obesity have higher rates of anxiety disorders, and GLP-1 receptors sit in brain regions that handle stress and reward as well as appetite. So it was always likely that someone would ask whether GLP-1 microdosing changes anxiety, for better or worse.
This article walks through the evidence that exists: randomised trials pooled into meta-analyses, national registry studies, large health-record cohorts, and the FDA's adverse event database. All of it used prescribed doses. None of it tested a sub-label microdose, and we flag that gap wherever it matters.
Key takeaways
- •A 2026 meta-analysis of 86 randomised trials and 133,378 participants found no difference in anxiety between GLP-1 medicines and control (risk ratio 0.92).
- •In a Swedish national cohort of 95,490 people with depression or anxiety, semaglutide was linked with a lower risk of worsening anxiety (adjusted hazard ratio 0.62).
- •Panic-attack reports for semaglutide in the FDA database rose from 6.7 per 10,000 reports in 2018-2022 to 42.0 in 2023-2024, a pattern researchers described as consistent with notoriety bias.
- •Health-record studies comparing GLP-1 users with users of other diabetes drugs found small differences in anxiety risk, with hazard ratios of 1.07 and 1.09.
- •No study has measured anxiety at sub-label microdoses, so effects at low doses are not established.
Can GLP-1 microdosing cause anxiety?
The randomised trials do not show GLP-1 medicines causing more anxiety than placebo or other treatments. Han and colleagues (Diabetes, Obesity and Metabolism, 2026) pooled 86 randomised trials with 133,378 participants. Psychiatric disorders as a group were no more common on a GLP-1 (risk ratio 0.96), and the same held for anxiety, depression, suicide and sleep disorders.
Psychiatric adverse events in randomised GLP-1 trials (risk ratio vs control)
Han S et al., Diabetes Obes Metab 2026, meta-analysis of 86 RCTs, 133,378 participants. A value of 1 means the same rate as control; none of these differences were statistically significant.
The authors checked whether the answer changed by drug, dose, treatment length or baseline BMI, and it did not. They also tested whether bigger drops in weight or blood sugar moved psychiatric risk, and found no effect there either.
Trial data have one weakness for this question. Anxiety was counted only when it was reported as an adverse event, so mild worry that never reached a doctor would be missed. The trials also tended to exclude people with unstable mental illness. The meta-analysis tells us GLP-1 medicines do not produce a broad rise in anxiety. A trial result like this cannot rule out a small group of people who feel worse.
What did the liraglutide weight-loss trials find?
The liraglutide weight-management trials found anxiety at nearly the same low rate as placebo. O'Neil and colleagues (Diabetes, Obesity and Metabolism, 2017) pooled the SCALE trials, in which 3,384 people took liraglutide 3.0 mg and 1,941 took placebo, alongside a reduced-calorie diet and exercise.
Anxiety and depression events in the SCALE trials (events per 100 person-years)
O'Neil PM et al., Diabetes Obes Metab 2017. Pooled phase 2 and 3a trials, liraglutide 3.0 mg vs placebo. Events captured through adverse event reporting.
- Liraglutide 3.0 mg
- Placebo
These trials also handed out mental-health questionnaires at set visits, which catches symptoms that patients might not volunteer. Depression scores on the PHQ-9 fell slightly in both groups, from 2.8 to 1.8 on liraglutide and from 2.9 to 1.9 on placebo. The baseline scores were low, which reflects a trial population without much active mood illness.
Does semaglutide help people who already have anxiety?
Semaglutide was linked with less worsening of anxiety in a large Swedish study, though that is not proof that it treats anxiety. Taipale and colleagues (The Lancet Psychiatry, 2026) followed 95,490 people with a diagnosis of depression or anxiety who used diabetes medicines between 2009 and 2022. Of them, 22,480 used a GLP-1 at some point.
The study compared each person with themselves, looking at periods on a drug against periods off it. That design removes stable differences between people, such as genetics or lifelong personality. "Worsening mental illness" meant a psychiatric hospital stay, more than 14 days of psychiatric sick leave, a hospital stay for self-harm, or suicide.
Risk of worsening mental illness by GLP-1 drug (adjusted hazard ratio vs non-use)
Taipale H et al., Lancet Psychiatry 2026. Swedish national registers, people with depression or anxiety. Within-individual design. Below 1 means lower risk while on the drug.
Semaglutide stood out. Its adjusted hazard ratio for worsening anxiety on its own was 0.62, and for worsening depression 0.56. Liraglutide was linked only with less worsening depression. Exenatide and dulaglutide showed no difference. GLP-1 drugs as a group were linked with fewer self-harm hospital stays (adjusted hazard ratio 0.56).
The authors suggested semaglutide and liraglutide might serve as dual-purpose options for people with diabetes or obesity and mood problems, and called for randomised trials to test it. A Taiwanese cohort by Tsai and colleagues (Frontiers in Pharmacology, 2022) pointed the same way. Among 10,690 GLP-1 users with diabetes, combined anxiety or depression ran at 6.80 per 1,000 person-years, against 9.36 in 42,766 matched non-users. Use for longer than 180 days was linked with an anxiety rate of 2.93 per 1,000 person-years. For more on mood, see our article on GLP-1 microdosing, mood and depression.
Why do some studies find more anxiety on GLP-1 medicines?
Studies that find more anxiety usually compare GLP-1 users with people who took no weight or diabetes medicine at all, and those groups are not alike. Kornelius and colleagues (Scientific Reports, 2024) matched 162,253 patients with obesity and reported a 108% higher risk of anxiety and a 195% higher risk of major depression among GLP-1 users. People who seek out a weight medicine may already be under more strain, see doctors more often, and get diagnosed more often.
Comparing GLP-1 users with people starting a different diabetes drug narrows that gap. Anagnostakis and colleagues (Diabetes Research and Clinical Practice, 2026) studied metformin users with type 2 diabetes in TriNetX records over a median of 5.5 years. Anxiety disorder risk was slightly higher on a GLP-1 than on an SGLT2 inhibitor (hazard ratio 1.09) or a DPP-4 inhibitor (1.07), and no different from a sulfonylurea.
Anxiety risk in three cohort comparisons (hazard ratio)
Taipale H et al., Lancet Psychiatry 2026 (worsening anxiety in people already diagnosed, within-individual). Anagnostakis F et al., Diabetes Res Clin Pract 2026 (new anxiety disorder diagnoses, active comparators). Different outcomes and populations, shown together to illustrate how design changes the answer.
A 2026 systematic review by Yi and colleagues (Psychoneuroendocrinology) summed it up. Animal studies consistently showed less anxiety-like behaviour on GLP-1 drugs, while human cohorts were mixed but leaned toward lower anxiety. Randomised trials designed around anxiety itself have not been done.
Does Ozempic cause panic attacks?
Current evidence does not show that semaglutide causes panic attacks, though a small effect has not been ruled out. Melamed and colleagues (Pharmacoepidemiology and Drug Safety, 2026) searched the FDA Adverse Event Reporting System. Panic attacks appeared in 87 of 33,267 semaglutide reports (0.26%), compared with 0.05% for both dulaglutide and empagliflozin.
That gap shrank once the analysis adjusted for age, sex and the reason for treatment, to reporting odds ratios of 2.82 and 2.75. Restricted to reports filed before July 1, 2023, the adjusted figures were 1.38 and 1.14, and neither was statistically significant. The surge came afterwards, when GLP-1 mental-health stories and regulatory reviews were in the news.
Panic-attack reports per 10,000 semaglutide reports in FAERS
Melamed B et al., Pharmacoepidemiol Drug Saf 2026. Spontaneous reports cannot measure how often an effect occurs.
The researchers called this pattern notoriety bias: once people hear a drug might cause something, they report it more. They also reviewed 12 studies of semaglutide, including randomised trials and active-comparator cohorts, and found no consistent increase in panic attacks or acute anxiety. Their conclusion was careful. The data neither prove nor exclude an effect in some individuals.
Does a lower dose mean less anxiety risk?
Nothing shows that a lower dose is gentler on anxiety, and one study points the other way. In the trial meta-analysis, dose made no difference to psychiatric outcomes. Murugadoss and colleagues (npj Metabolic Health and Disease, 2026) followed 63,215 semaglutide users who already had neuropsychiatric conditions. Higher attained doses were linked with lower later rates of anxiety- and stress-related disorders.
That study is observational, and people who reach higher doses may simply be the ones who tolerate the drug well. It does undercut one common assumption among microdosers, that a smaller dose is automatically the safer choice for mood. At sub-label doses, the anxiety effect is simply unknown.
Can stopping a GLP-1 make anxiety worse?
One study suggests stopping may matter more than starting. Zhang and colleagues (Nature Metabolism, 2026) used Shanghai hospital records for people with type 2 diabetes. While on treatment, GLP-1 users had a neutral risk of new anxiety or depression compared with DPP-4 inhibitor users, and a lower risk than SGLT2 inhibitor users. After they stopped, their risk was higher than for people who had stopped either comparison drug.
Part of that rise tracked with higher triglycerides after stopping, which hints at a metabolic link. The finding is new and comes from one population. Even so, it is one more reason to plan breaks and stops with a prescriber. Our guide to missed doses and restarting after a break covers the practical side.
What should you do if you feel anxious on a GLP-1 microdose?
Tell the prescriber and track the pattern. The first weeks on a GLP-1 bring less food, sometimes less sleep, and sometimes nausea, all of which can feel like nerves. Caffeine on an emptier stomach can add to it. Our piece on coffee and caffeine on a microdose looks at timing.
Useful details to bring to the appointment:
- •When the anxiety started relative to the first dose or the last dose change
- •Whether it peaks on injection day or the day after
- •How much you are eating and drinking, and how you are sleeping
- •Any past history of anxiety, panic or depression, and current medicines for it
Anyone having thoughts of self-harm should get help the same day, through a doctor, an emergency department or a crisis line. Do not stop or change a prescribed psychiatric medicine because a GLP-1 was started without talking it through first.
Randomised trials show no overall rise in anxiety on GLP-1 medicines, and some large cohorts link semaglutide with less worsening anxiety. The panic-attack signal looks driven by reporting. None of this was measured at microdoses, so individual reactions still deserve attention.
Frequently asked questions
Can a GLP-1 microdose cause anxiety?
No study has tested anxiety at sub-label microdoses, so the honest answer is that it is unknown. At prescribed doses, a 2026 meta-analysis of 86 randomised trials found no difference in anxiety between GLP-1 medicines and control (risk ratio 0.92). Individual people can still feel unsettled in the first weeks, and that is worth raising with the prescriber.
Does semaglutide help with anxiety?
It has not been proven to. A Swedish national cohort of people already diagnosed with depression or anxiety linked semaglutide with a lower risk of worsening anxiety (adjusted hazard ratio 0.62) compared with periods off the drug. That is an observational finding, and the authors called for randomised trials before drawing conclusions.
Does Ozempic cause panic attacks?
The evidence does not show a consistent link. Panic attacks were reported more often for semaglutide than comparators in the FDA's adverse event database, but the signal shrank after adjustment and was not significant before July 2023. Reports per 10,000 rose from 6.7 in 2018-2022 to 42.0 in 2023-2024, a period of heavy media attention, a pattern the authors described as consistent with notoriety bias.
Why do some studies find more anxiety on GLP-1 medicines?
Mostly because of who is compared with whom. A cohort of 162,253 matched patients with obesity found a 108% higher anxiety risk in GLP-1 users versus non-users, but people who start a weight medicine differ from people who do not. Studies that compared GLP-1 users with users of other diabetes drugs found much smaller differences, with hazard ratios of 1.07 and 1.09.
Can stopping a GLP-1 make anxiety worse?
One study suggests it might. In Shanghai health records, people with type 2 diabetes had a neutral anxiety and depression risk while on a GLP-1 compared with DPP-4 inhibitors, but a higher risk after they stopped. The finding has not been confirmed elsewhere, and it is a reason to plan any stop with a prescriber rather than quitting abruptly.
Is a lower GLP-1 dose safer for anxiety?
There is no evidence that it is. The trial meta-analysis found no difference in psychiatric outcomes by dose. An observational study of 63,215 semaglutide users with preexisting neuropsychiatric conditions found the opposite direction: higher attained doses were linked with lower later rates of anxiety- and stress-related disorders, though that design cannot prove cause.
Should people with an anxiety disorder avoid GLP-1 medicines?
Current evidence does not say they must. The Swedish cohort studied 95,490 people with depression or anxiety and found semaglutide and liraglutide linked with less worsening of mental illness, not more. Anyone with a history of anxiety, panic or depression should still tell the prescriber, so that mood can be checked at follow-up visits.
What should I do if I feel anxious after starting a GLP-1?
Tell the prescriber, and describe when it started and how it relates to injection days, meals and caffeine. Low food intake, poor sleep and dehydration in the early weeks can all feel like anxiety. Any thoughts of self-harm need urgent help the same day through a doctor, an emergency service or a crisis line.
Clinical references
- Han S, Yang Y, Liu Z, et al. Association of Glucagon-Like Peptide-1 Receptor Agonist Use With Risk of Psychiatric Disorders: A Systematic Review and Meta-Analysis of Randomised Controlled Trials. Diabetes, Obesity and Metabolism. 2026. doi:10.1111/dom.70692
- Taipale H, Taylor M, Lähteenvuo M, et al. Association between GLP-1 receptor agonist use and worsening mental illness in people with depression and anxiety in Sweden: a national cohort study. The Lancet Psychiatry. 2026. doi:10.1016/S2215-0366(26)00014-3
- Melamed B, Parsi-Zahabi M, Omrani MA, et al. Semaglutide and Panic Attacks: Active-Comparator Disproportionality Analysis, Temporal Assessment of Notoriety Bias and Evidence Triangulation in Post-Marketing Pharmacovigilance. Pharmacoepidemiology and Drug Safety. 2026. doi:10.1002/pds.70517
- O'Neil PM, Aroda VR, Astrup A, et al. Neuropsychiatric safety with liraglutide 3.0 mg for weight management: Results from randomized controlled phase 2 and 3a trials. Diabetes, Obesity and Metabolism. 2017. doi:10.1111/dom.12963
- Tsai WH, Sung FC, Chiu LT, et al. Decreased Risk of Anxiety in Diabetic Patients Receiving Glucagon-like Peptide-1 Receptor Agonist: A Nationwide, Population-Based Cohort Study. Frontiers in Pharmacology. 2022. doi:10.3389/fphar.2022.765446
- Kornelius E, Huang JY, Lo SC, et al. The risk of depression, anxiety, and suicidal behavior in patients with obesity on glucagon like peptide-1 receptor agonist therapy. Scientific Reports. 2024. doi:10.1038/s41598-024-75965-2
- Anagnostakis F, Kokkorakis M, Anastasiou G, et al. Association of glucagon-like peptide-1 receptor agonist use with anxiety disorders, depression, self-harm, and suicidality: a large cohort study. Diabetes Research and Clinical Practice. 2026. doi:10.1016/j.diabres.2026.113104
- Zhang T, He P, Ji Y, et al. Association of GLP-1RA discontinuation and risk of depressive and anxiety disorders in people with type 2 diabetes: a cohort study. Nature Metabolism. 2026. doi:10.1038/s42255-026-01567-z
- Murugadoss K, Venkatakrishnan AJ, Soundararajan V. Higher semaglutide dose is associated with lower neuropsychiatric event incidence independent of weight loss. npj Metabolic Health and Disease. 2026. doi:10.1038/s44324-026-00131-3
- Yi YT, Zheng YJ, Bargiota SI, et al. The effect of GLP-1 receptor agonists on anxiety: A systematic review. Psychoneuroendocrinology. 2026. doi:10.1016/j.psyneuen.2026.107844
Medical disclaimer. This article is for general educational and informational purposes only. It is not medical advice and not a substitute for a clinician who knows your history. The studies described here used prescribed doses, most are observational, and sub-label "microdose" regimens have not been studied for anxiety or mood. GLP-1 medicines are not approved to treat anxiety. Talk to your prescriber before starting, changing or stopping any medicine, and seek urgent help for thoughts of self-harm. Charts show values reported in the cited sources.